Malan loess is widely distributed on the Chinese Loess Plateau and poses great challenges to geotechnical, ecological, and agricultural practices due to its unique structure and collapsibility. It is essential to understand the evolution of these properties with depth to assess soil stability and reduce engineering risks in the area. This study investigates the mechanical properties and microstructural evolution of Malan loess with depth and employs multivariate statistical methods to explore their complex interrelationships. Oedometer-collapse tests reveal a 94.2 % reduction in collapsibility coefficient (delta s) from 0.0722 at 1 m to 0.0042 at 9 m, indicating a significant reduction in collapsibility with increasing depth. According to the results of the direct shear test, it showed that the shear strength initially decreases and then increases due to the combined effect of the water content and dry density. Scanning electron microscopy (SEM) images reveal the densification of the loess structure, with changes in particle contact from point to face contact and the evolution from macropores to mesopores and small pores as depth increases. Quantitative analysis by Avzio showed a decrease of 61.5 % in macropores area and an increase of 62.5 % in small pores area. The results obtained by Pearson's correlation analysis and random forest model showed that among these microstructural characteristics, the total pore area (%IncMSE = 22.77 %) is the most important factor influencing the collapsibility properties of loess and water content (%IncMSE = 17.72 %) acts a key role in controlling shear strength. Additionally, compared to traditional methods, the random forest model offers a more insightful understanding of nonlinear relationships and multifactorial coupling effects. These findings provide scientific guidance for geotechnical engineering in loess regions, aiding in risk mitigation and promoting sustainable construction.

Subgrades constructed from loess-a loose and porous material-demonstrate significant compressibility and collapsibility. To study these properties of loess subgrades, this article proposes a vertical vibration compaction method (VVCM) that provides a reliable simulation of field compaction and investigates the factors influencing the deformation characteristics of loess subgrade by VVCM-prepared specimens. The results show that the correlation between the compression modulus of loess samples prepared by VVCM and that of core samples obtained from the construction site is more than 85 %. In addition, the deformation resistance of the VVCM sample is better than that of the traditional quasistatic compaction method (QSCM) sample. Under the same compaction factor and water content, the compressive modulus of VVCM sample is at least 10 % higher and its collapsibility coefficient is 10 % lower than that of QSCM sample. With the increase in compaction factor, the compression modulus increases and the collapsibility coefficient decreases, indicating improved resistance to compressive deformation and reduced susceptibility to collapse in loess. With the increase in water content, the compression modulus and collapsibility coefficient decrease, reflecting greater compressibility and increased collapse resistance in loess.

The evolution of loess microstructure exerts a direct impact on its collapse evolution during dry and wet (DW) cycles. In this study, a hydro-mechanical coupling numerical model considering DW cycles and mechanical loading was established by extending the Barcelona Basic model, meanwhile combining with the test results to reveal the effect of DW cycling on the collapse deformation and strength response of loess. Additionally, the microscopic mechanism of loess collapse evolution was revealed through microscopic tests. Results indicated DW cycles caused the net compaction of loess, with the first DW cycle exerting the most significant effect on its deformation, consequently deteriorating the loess. Wetting under constant loading leads to a collapse of macrostructures formed by aggregates. Moreover, DW cycles transformed the structural units from line and surface contact to point. The basic structural units exhibited obvious grade properties, in which DW cycles trigger the collapse of compound aggregates, with the number of relatively stable mononuclear aggregates and intergranular pores increasing. DW cycles in an open environment induced the loss of cementing materials such as soluble salts and reduced the bonding strength among basic structural units. This subsequently tended to weaken the structural properties of loess and decreased the mechanical properties.

Increasing extreme weather events and climate change can significantly affect soil moisture regimes, particularly soil suction, leading to additional challenges associated with unsaturated soils, including the collapse phenomenon. The collapsibility of soils poses significant engineering and geotechnical risks globally, necessitating urgent attention from engineers. This work establishes a numerical model of a shallow foundation subjected to rainfall and load using COMSOL Multiphysics. A hydromechanical model (H-M) is introduced which incorporates The Richards' module and the Extended Basic Barcelona Model (EBBM) as a constitutive model to predict settlements in shallow foundations influenced by climate change and intense rainfall. The validation of the model is conducted through experimental tests, ensuring its accuracy. Additionally, in the practical application, the hydromechanical model is applied to anticipate the effect of infiltration on settlements of shallow foundations. The simulation results show that infiltration leads to an increase in the pressure head above the water table, decreasing soil suction, which induces additional settlement due to wetting-induced collapse. The maximum settlement happened at the corners of the footing due to increased exposure to infiltration and a greater reduction in suction. The collapse potential calculated from the numerical simulation was found to be consistent with the predictions established via analytical models, validating the accuracy of the numerical approach.

Loess tunnels are very common in the Loess Plateau, and they pose unique geological threats. Loess tunnels are often difficult to detect and control due to their concealment and sudden appearance. Thus far, research on the genesis and evolution of loess tunnels remains scarce. In this paper, the genesis and evolution mechanism of the loess tunnels in the Loess Plateau is studied in depth, and the location, shape, and size information are obtained via field investigations. The potential correlations between the loess sediment, the basic physical properties (depth, water content, particle size composition, collapsibility coefficient, and self-weight collapsibility coefficient), and the tunnel density are inferred based on the Pearson's correlation coefficients and tests on the physical and mechanical properties of the loess sediments. In addition, spatial statistical modelling is employed to justify and predict the observed spatial distribution of the loess tunnels assuming Gaussian Markov random fields. The formation of loess tunnels is due to a combination of factors, including the formation thickness, soil properties, joints and fissures, topography, hydrogeology, and climatic conditions. The thickness of the loess, loess sediment properties, and their spatial relationship jointly determine the material basis of the formation of the loess tunnels. The loess tunnels at different depths have different main controlling factors that are hierarchical by depth. The evolution process of loess tunnels can be divided into five stages: the incubation stage, formation stage, development stage, failure stage, and withered stage. The characteristics of each stage are discussed in detail. Our work provides novel insights into subsurface erosion from the aspect of soil tunnels. It improves our understanding of hill slope geomorphological evolution and also provides effective techniques for tunnel erosion control.

Loess is a problematic type of soil with a worldwide distribution due to its collapsibility. The temporal discontinuity and spatial nonuniformity of its collapsibility can bring severe damage to building foundations, roads and water pipelines. In this study, the relationship between the saturation and K-G model parameters is established based on indoor compression tests and collapsible tests; the deformation characteristics of loess immersed in water are studied via a large-scale trial immersion pit test. The test site is a circular pit with a diameter of 10 m. The loess is immersed for 46 days; the variation in its accumulated settlement over time is recorded for 60 days, and its deformation process is simulated using a self-designed programme. Results show that for the stress-strain relationship of unsaturated loess, the relationship between equivalent suction and saturation can be obtained through the principle of deformation equivalence and fitted using the exponential function. The maximum vertical displacements calculated in the simulation and on-site immersion pit experiment are 0.036 m and 0.032 m, respectively. Such relatively good consistency indicates that the proposed method can reasonably predict the collapse behaviour of loess due to immersion. This research provides a reliable method for the numerical simulation of loess immersion deformation, and the parameters in the model only need to be determined by conventional experiments.

Loess is widely distributed in China and it is commonly considered as the problematic soil due to its collapsibility subjected to the water invasion. The microstructure plays an important role in the mechanical properties of the loess soil. In this note, the microstructures of intact loess samples and the inundated loess sample were investigated by using both mercury intrusion porosimetry (MIP) and nuclear magnetic resonance cryoporometry (NMRC). It is observed from the results of both MIP and NMRC tests that the intact loess has a multi-model pore size distribution function while the inundated loess has a unimodal pore size distribution function. As the coefficient of collapsibility (delta s) is a key parameter commonly used for the evaluation of the engineering properties of the loess, the delta s of the specimens tested under different conditions was measured. Subsequently, a new multi-variable linear model was proposed for the estimation of delta s from the index properties based on the results of factor analyses. The estimated results of delta s from the proposed model show good agreements with the measured data.

The Guanzhong Plain city group is located in the centre of inland China, which is an important fulcrum of the Asia-Europe Continental Bridge, and also the second largest urban agglomeration in the western region of China. However, the large thickness collapsible loess has always been a pain point in the process of urban development. In this paper, for the problem of collapsible loess in Guanzhong area, five typical geomorphological units in Xi'an, Xianyang, Tongchuan and Sanmenxia are targeted for sampling. Through a large number of indoor and outdoor experiments, the relevant data of physical properties, hydrophysical properties, mechanical properties, collapsible index of collapsible loess and the distribution pattern of each index in the study area are analysed.By analyzing the correlation among the indexes, four indexes which are closely related to the collapsibility of loess are obtained: natural density, void ratio, dry density and saturation, and the regression analysis of these four indexes and the collapsibility coefficient is carried out respectively.This finding provides a reference for a more in-depth study of the characteristics of collapsible loess in Guanzhong area, and can optimise the engineering investigation process in urban construction, which is very important for the urban construction in this area.